摘要:Robust portfolio optimization theory is an essential foundation for modern financial modeling, which is awell-studied but not fully conquered territory. Conservatismis one of the most discussed issues by numerous scholars.To obtain a robust portfolio model with satisfactory performance, we propose the hybrid robust mean-variance portfoliomodel constrained with different ellipsoidal uncertainty setsin this paper. Additionally, skewness is also considered inthe objective function. Preselection is designed for pickingout the high-quality risky assets, where two machine learningalgorithms, Random Forest and Support Vector Machine, areinvolved. In the numerical experiments, the US 48 industrydata set from Kenneth R. French is employed to verify theeffectiveness of the proposed hybrid portfolio models. Thecomparative results between the proposed hybrid modelsand baseline portfolio models (equal-weighted model, meanvariance model, mean-variance-skewness model) show thatthe proposed hybrid robust mean-variance portfolios considering skewness with preselection beat the baseline strategiesby a clear margin. Also, the actual effectiveness of skewnessin the hybrid robust models is analyzed.